Colorectal cancer, as the second leading cause of cancer-related mortality in the USA, is a major public health problem. This proposal plans to investigate the chief secretory product of colonic epithelium:mucin glycoproteins. Mucins are large glycoproteins consisting of carbohydrate side chains on apomucin protein backbones. It has long been known that oligosaccharide side chains of colon cancer mucin carry different ("cancer-associated") epitopes than the oligosaccharides of normal colonic mucin. It is not clear whether this abnormal mucin phenotype of colon cancers contributes to the biological behavior of the tumor, and if so, what mechanism(s) accounts for the aberrant glycosylation. Our previous studies disclosed that the sialosyl-Tn antigen (STn), a mucin-associated disaccharide structure, is not expressed by normal colonic epithelium but becomes exposed in the vast majority of colon cancers. Moreover, expression of STn by colon cancer tissues was an independent predictor of a poor prognosis. This suggests that colonic mucin expressing STn may in fact contribute to the biological behavior of the cancer cell. It also supports earlier histopathological observations that highly mucinous colon cancers are associated with a poor prognosis, and recent experimental models showing that mucin contributes to the metastatic potential of the colon cancer cell. Since mucin, and specific mucin carbohydrates, appear to contribute to colon cancer behavior, we have been exploring the mechanisms involved in mucin gene-specific glycosylation and mucin gene expression. During the last project period we developed clonal cell lines from human and rat colon cancers that differ in STn expression. Preliminary characterization of the human clones indicates that one synthesizes only short oligosaccharides (Tn.STn) whereas the other expresses only longer structures. These cells will now serve as excellent models for dissecting mechanisms involved in glycosylation. Characterization of the rat clones will be useful for subsequent in vivo studies of colon cancer cell behavior and immune response. We have also begun to explore the repertoire of mucin gene expression in the colon, finding that MUC1 and MUC2 genes are often reciprocally expressed, and MUC2 and MUC3 genes are often decreased in colon cancers. The present project has three aims: 1) To define the molecular mechanisms responsible for STn+ and STN- phenotypes in model cell lines; 2) To use these model cell lines to elucidate mucin-type specific and organ-specific glycosylation employing novel mini-mucin gene constructs; 3) To further define the repertoire and pathological significance of mucin gene expression in malignant and premalignant colonic tissues. These studies should provide direct insight into mucin glycosylation processes in the colon and will lay the groundwork for studies on mechanisms of mucin secretion, and the role of mucin in immune response to colon cancer.